Field Classification of of Soils with < 25% Fines 3

[Matsuri]

We have some lab results that classify soil as SC without corresponding Atterburg limits. The sieve analysis show less than 25% passing a #200 sieve. When questioned how the classification was reached the respons was it was based on field observation. Is it reasonable or possible to discern between an SC and an SM soil with this percentage of fines by visual observation of samples during collection or is lab testing required?

Thanks,

 

[Matsuri]

To Risk Taker,
Soil will be placed as backfill at 95% modified proctor starting well below the water table up to final grade.

 

[msucog]

in general i wouldn't expect liquefaction potential to be high or even moderate for rhis fill material placed to 95% modified proctor compaction. natural soils would possibly be more of concern. also, as has been presented before (can't remember the author off the top of my head), fine grained soils may still possess liquefaction potential so don't rule it out just based solely on fines content...but well compacted fill pretty well knocks it out due to the reduction in void ratio. but as a geotech i always get clarification on those "bad" specs before playing cowboy on my own due to the added headaches if i simply disregard the specs (regardles of how right or wrong they are). often a simple phone call and conversation with the team can clear that up.

 

[dgillette]

Plastic soils generally ARE NOT liquefiable due to their plasticity. There are exceptions, with low PI and high void ratio (much looser than 95% of modified!!), although sometimes it's hard to tell liquefaction from sensitivity. Assuming there is enough fines that they dominate the behavior, water content of the minus #40 fraction has to be more than 80 - 85% of LL, and PI has to be less than 12 - 18%, according to Bray and Sancio in the September 2006 ASCE JGE. (The ranges shown here are the "maybe" range, where testing is required.) Idriss and Boulanger's red book looks at a little differently, but you would get to the same place by their approach.

The original "Chinese criteria" (less than 15% smaller than 5 microns, LL<35, %w> 0.9 LL) have been shown by a number of researchers not to work, in part because of the limited data set used by Wang to develop them. Likewise, there is no fines content that can rule out liquefaction; there are a number of well-known cases where ~100% fines liquefied (Moss Landing, a loess deposit in Tajikistan, tailings, etc.).

95% of modified would be similar to or higher than what we use for large dams, where we typically require 98% of standard Proctor. Liquefaction can be ruled out DEFINITIVELY with that kind of density!

 

[msucog]

right on dgillette!
(and sancio is the author i was trying to think of)

 

[RiskTaker]

I agree. With that density, liquefication should not take place. But here in Texas the clays I have came across have a possibility of shrink/swell when water is added.

 

[BigH]

dgillette - agree about not "liquefying" but what about loss of shear strength? If the threshold pore pressure is exceeded, there could be a loss of shear strength as pore pressurs "build up" and hence potential shear distress/failures (not liquefaction - but a direct consequence anyway).

 

[dgillette]

Big H - You're referring to typical clays, NOT those at 95% of modified, right? At 95% of modified, the material probably needs 15 m or more of overburden pressure to make it be contractive. We don't worry much about that stuff too much.

But yes, you are correct to pay attention to decrease in strength w/ cyclic loadings if they will push the peak strength. In general, saturated clays show sensitivity, i.e., remolded strength less than peak, and will show decrease in strength with very large cyclic shear strains and buildup of excess PWP, but just loss of stiffness with smaller cyclic shear stresses that stay below the peak strength. Sensitivity ratio can be anywhere from 1.5 to VERY high in Scandinavian and Quebec quick clays, with 2 to 4 being most common. Idriss and Boulanger recently ~2007 published a big report on cyclic failure of clays, sensitive and otherwise. (I'm fairly sure that it's not in the public domain, so I can't post it for you, unlike that report I promised to post for McCoy last week.) Like their liquefaction red book, it will probably be considered the state of practice in the near future. They show cyclic failure eventually occurring with repeated loads that flirt with the monotonic peak strength but don't quite reach it (like 95%).

Funny you should bring this topic up. In a dynamic deformation analysis, the trick (OK, one of many tricks) is deciding how much deformation it takes to reduce the strength from peak to fully softened to remolded strength. I was considering posting a new question to see if anybody knew of any well analyzed case histories or centrifuge research or anything else that would provide a good analog. The one good case history I know of is the 4th Ave, Anchorage AK slide in 1964, on which I have papers by Idriss and by Stark and Contreras, and of course that is different geology, stress history, clay properties, etc. from the case I'm interested in.

Every time I turn around, this s(tuff) gets more complicated.

DRG

 

[dgillette]

Son of a gun. Here it is:

http://cee.engr.ucdavis.edu/faculty/boulanger/PDFs/2004/Boulanger_Idriss_CGM04-01_2004.pdf

 

[BigH]

Thanks Dave - and you all have a wonderful Holiday . . .